The second floor gallery of the Wormfarm Institute in Reedsburg, Wisconsin, is a far cry from the funky glassware and biosafety protocols of a working microbiology lab. One corner contains an improv kitchenette, circa 1987. Sprout — the mascot offspring of the Jolly Green Giant — waves improbably from behind a cubicle wall across the room; the child-size plastic statue is missing a hand. A cool draft of autumn rain and small-town traffic noise flows through an open window.
Assistant professor of bacteriology Federico Rey chats with his Wormfarm hosts as more than three dozen attendees of the annual Fermentation Fest assemble into a casual arc of folding chairs and a couple of vintage couches. These are people who already understand the idea of microbes as friends. Makers of coleslaw and kimchi, kombucha and beer — they are motivated by microbes and have paid to hear Rey’s summary of the state of current human microbiome research.
Across the life sciences, the microbiome is the buzziest of buzzwords, invoking a symphony of hope, hyperbole, and high expectations. Rey shares in the overall enthusiasm, but he is careful about the speculative details. Yes, the microbiome might even match our frothiest expectations. And no, he can’t cure your diabetes or make you leaner, faster, or smarter. He can’t even tell you if your microbiome is healthy. Not today.
Because nobody can.
In front of a large wall hanging of textile orange circles representing the bubbles of fermentation, he begins where he has to begin, very near the beginning. “Microbes are the most abundant form of life on this planet,” he says, his thick Argentinian accent backlit by a docent’s enthusiasm. “They can live in places where we cannot imagine life.”
“Microbes outnumber us by many orders of magnitude. They power almost everything on the earth,” he says. They convert as much carbon dioxide into organic compounds as plants do and emit more methane than the oil and gas industry. “Literally, there is no place on earth where there are no microbes,” he continues.”It is impossible to get rid of them.” And despite our germophobia, they’re mostly good company: “A very small fraction of microbes are pathogens. Most of them are commensal — they don’t do good or bad — or they are beneficial for humans.” These last ones, the microbes that fuel our fantasies of easy cures and everlasting health, truly capture Rey’s interest.
Every surface on our bodies is colonized by some kind of microbe, and microbiologists have identified thousands of species of bacteria that can inhabit the human gut. Each of us, in turn, has a collection of between 100 and 200 different bacteria strains, comingled with other life forms from fungi and protists to viruses and archaea. These enteric ecosystems — different for every single human, even more unique than a fingerprint — each contain 100 times the genetic information of our own cells. They both supplement and interact with our bodily blueprints.
Deciphering who’s who is not even half the problem. How exactly do our bodies gather these microbes? What shapes the resulting ecosystem? How do humans and microbes interact? In 2017 alone, Rey published new research about microbiome effects in diabetes, Alzheimer’s, and the cycling of the nutrient choline (which may positively affect fetal brain development but also can lead to heart disease later). “Every single disease or health condition scientists look at, they find a microbiome connection,” he says.
And yet there is no single definition of a healthy microbiota. And what is healthy for you may not be healthy for me.
Federico Rey arrived in the United States in 1999 with advanced biological questions on his mind and little idea that microbes could hold the answer. As a research fellow at the Henry Ford Health Sciences Center in Detroit, he focused on hypertension and vascular disease. But when he moved to the University of Iowa for his Ph.D. in 2001, he met Rhodopseudomonas palustris during a lab rotation. These extravagantly versatile bacteria are known for their ability to use four different modes of metabolism to scavenge energy, nitrogen, and carbon from a variety of sources — with or without oxygen. Intrigued by the diversity and adaptability of microbes, Rey says he fell in love with the bacterium.
R. palustris was an obvious stepping-stone towards biofuels. But in 2005, Rey saw a talk by Jeff Gordon, a pioneer in the study of human-microbiome interactions. Gordon began examining the development of the mammalian gut in the 1980s. Eventually he realized that microbes were essential to the process, and he set out to untangle this complex relationship using early sequencing techniques and transgenic and germ-free mice.
Rey joined Gordon’s lab at Washington University in St. Louis as a postdoc in 2006 just as microbiome science was gaining momentum. Tools for reading the genetic code were getting faster, cheaper, and more versatile. Computers used to crunch burgeoning data sets were growing in power as new statistical methods were increasing in sophistication.
One of the earliest hints of the power of microbes was that transplanting the microbiome from one mouse to another could also transfer basic metabolic conditions, such as obesity. It was in Gordon’s lab that Rey first met CALS professor of biochemistry Alan Attie — through the microbes of his mice. In Attie’s efforts to unravel the many mysteries of diabetes, his lab sent Gordon microbiome samples from genetically distinct mice that had been placed on a high-fat or a chow (i.e., grain-based) diet. It was known that both diet and genetics had a significant effect on the metabolic health of these mice. Gordon helped to show that the microbes played a role as well.
Rey took the project with him when he was hired by CALS in 2013, and he’s been collaborating with Attie since. It’s a task of daunting complexity: integrate two genetically complex systems that play a role in metabolic disease. Neither is close to being perfectly understood, yet they interact with each other, coalescing in each individual.
Hundreds of mammalian genes are already understood as part of the metabolic pathways that go awry on the way to diabetes. The human microbiome, meanwhile, produces thousands of chemicals that act within the genetic framework of humans. We’ve long understood how these microbes help us break down the complex compounds in the plant-based foods we eat, providing about 10 percent more energy. More recently, we’ve learned how these microbial bioreactors produce molecules called short-chain fatty acids — acetic acid, propionic acid, butyrate, and other products of fermentation that signal our bodies in health-promoting ways.
Our bodies sense these molecules, helping regulate things like gastrointestinal motility. Less well understood are the thousands of chemicals that train our immune system and help regulate everything from kidney function to brain chemistry.
“This is one of the most important aspects of the microbiome that is revolutionizing biology,” Rey says. But unraveling the interplay between host genetics and metabolism is anything but easy.
In work published in Cell Reports in 2017, Rey’s student Julia Kreznar used eight different mouse strains and microbial transplants to help unravel this tangled web. The study found measurable differences in the microbiome established in the different strains. These microbes, in turn, influenced the likelihood that the mice would succumb to metabolic disease. The work also demonstrated a novel link between the gut microbiome and insulin secretion.
“We can show that microbiota affected pancreatic islet physiology and function,” Kreznar says. It’s a promising step, but it’s also an indirect interaction. Identifying the mediators of these microbe-host interactions is really challenging.
“It’s the dawn of a new field,” Attie says. “We have 3 or 4 pounds of organisms that are producing so many molecules, and we don’t know the least of them.”
“I’m feeling daunted,” he admits. “We had this idea that we would potentially connect the dots genetically. It is enormously complex, and it’s hard. It’s harder than we even thought it was going to be.”
A Big Genetic Black Box
Back in Reedsburg, Rey knows the probiotic question is coming, so he makes a preemptive strike: “When I talk about microbiota, I’m not talking about probiotics. The probiotics that you can get at Walgreens are basically microbes from dairy, microbes that can live in milk. Microbes that are not adapted to live in our intestines.”
If the microbiome has taught him anything, it’s that generalizations are tricky. “If they work for you, you should continue,” he emphasizes, trying to claim middle ground. Because of course, probiotics are microbes, and there actually is a lot of evidence that some may provide immune benefits.
But you won’t get a better microbiota by eating probiotics, and if you don’t eat your yogurt on Saturday, by Sunday those yogurt microbes are pretty much gone. They just pass through. “The effect is good as long as you keep eating them. And that is the perfect model for a company, right?” he says, with a tone of innocent mischief.
But if you look at probiotics through the lens of the microbiota, you have to acknowledge this essential truth: It’s been a landmark decade, but we still don’t have tools that are sophisticated enough to measure the microbiota in sufficient detail. We still don’t have an adequate biological understanding of what makes a healthy microbiota. And we barely understand the complex dance these microbes do with our body. Add in the fact that probiotics are dietary supplements, and thus not well regulated.
“You have to be careful. You have to do your research,” Rey says. “There are many different strains of probiotics, and there are big differences between different strains.” This is further complicated by the difficulty of even properly identifying bacteria, which can evolve rapidly. “There is a big difference between George Clooney the actor and somebody named George Clooney who lives in Atlanta. They have the same name, but they are completely different people,” he explains. “There are thousands of strains of Lactobacillus rhamnosus. Some may have an effect. Many of them probably don’t.”
In fact, it is our ability to identify these subtle differences in microbes that sparked the current revolution in microbiome science. First came 16S ribosomal RNA. Essential to the construction of proteins, 16S changes very slowly, which allowed scientists to, finally, reliably identify the members in a microbial community.
We talk about DNA as the book of life. At first, just reading a few pages was a chore. Then we developed machines to read more pages, faster. Then we learned how to read the sequels: DNA makes RNA, which makes proteins and enzymes responsible for carrying out cellular processes, including the making and breaking of sugar for energy. The technology responsible for reading these interrelated genetic codes is called next-generation sequencing, and it has powered this first golden age of microbiome research. The challenge now is sifting through that data to find biological meaning. Postdoc Lindsay Traeger PhD’15 is one of Rey’s primary number crunchers. “I’m attracted to very broad questions that we can throw a data hammer at and see what falls out,” she says.
Right now Traeger is focused on the next stage of collaboration with Attie and several other UW–Madison faculty including Joshua Coon (biomolecular chemistry), Karl Broman (biostatistics and medical informatics), and Brian Yandell (statistics). “We know that the gut microbiome is influenced by diet,” Traeger says. “But there is also this genetic component, which is a black box.”
Using genetically distinct mouse strains is advantageous when you’re trying to model a particular disease. But if you’re trying to tease out broader biological principles, using a single strain of mice could lead to bias. That’s why the labs are using a special breed called Diversity Outbred, a strategic genetic mash-up of common lab strains and some wild strains.
In one hand, Traeger has the genetic code of each individual mouse. And in the other, she has the genetic code of the microbiome of each mouse. Using advanced statistics, she’s searching for patterns that suggest some molecular matchmaking.
“I’m trying to identify how the host is selecting for or deselecting for the presence and abundance of certain microbial functions. Because the microbes are interacting somehow with the host.” One gene of interest is responsible for creating the important immune protein TNF-alpha, which plays roles in cancer and autoimmune disease. Early returns suggest that the TNF gene is also involved in sensing and responding to bacteria that have flagella.
Of course, the TNF gene is only one of about 23,000 genes, while the genes associated with bacterial flagella are just a few out of potentially hundreds of thousands. With numbers that big, there’s a lot of noise to filter out. And lots of distractions. “It’s a little hard to focus,” she laughs. “I think I could just spiral off. [Rey] keeps me thinking about the biology.” Rey’s endless creative energy helps. “He’ll just bust into the office and say, ‘I have this idea!’”
The immensity of the black box also keeps the work exciting. “We do think we’re going to find some interesting examples of interaction of host and microbiome,” she says.
Two T-Bones a Day
The microbiome is a dynamic force. Change the diets of lab mice, and overnight the communities that live in their gut change dramatically. So why care what microbes are in your gut if you can switch it up that quickly?
Rey points to his classic Argentinian upbringing as an example. “I grew up eating a T-bone for lunch and a T-bone for dinner,” he says. “Twenty-five years. And I miss it very much,” he quips, evoking another laugh from his audience. “But if I became vegetarian long term, I would definitely select for different microbial communities. And I would likely have new microbes colonize me.”
But even while your gut community is adaptable, the microbes in your gut can also have long-term consequences. That T-bone? “There are components in meat that microbes love and that cause problems,” Rey warns. “But you might not have them.” Which could mean several things: You’ve never been exposed to them, or you’ve been exposed but they didn’t take. Or maybe they’re there, but other microbes keep them in check. All of those are open biological questions, which now makes nutrition even more complicated.
In 2011, Stanley Hazen of the Cleveland Clinic published a paper linking microbes to the breakdown of lipid phosphatidylcholine (lecithin) into several choline-related compounds, particularly TMAO (trimethylamine N-oxide), a chemical already found to be a strong predictor of heart disease risk. Specifically, microbes metabolized choline into trimethylamine (TMA), which is then converted in the liver to TMAO.
While diet is a big part of this risk — eggs, milk, liver, red meat, poultry, shellfish, and fish are major dietary sources — the combination of microbial and host biology leading to TMAO accumulation intrigued Kymberleigh Romano PhD’17, who decided to dig deeper for her doctoral work with Rey and was co-mentored by Daniel AmadorNoguez, an assistant professor in bacteriology with expertise in metabolomics. Despite years of lab experience, she’d never worked with lab animals before, but she knew the time had come. “A lot of the phenotypes we study exist only in the context of a host-microbe interaction,” she says. “A test tube is never going to develop heart disease.”
First she needed microbes. Harvard researcher Emily Balskus had identified the genes involved in microbial conversion of choline to TMA, and Romano began looking for them in human-associated microbes and constructing experimental mixes of microbes. As she tinkered she found that, in mice at least, you need a TMA producer present in the intestinal microbiota to see TMA accumulation. And as you add more TMA-producing species, less choline is left for the host.
Even though she’d narrowed down the difference to a single organism in her custom microbiota, it still wasn’t enough. One bacterium contains anywhere from 3,000 to 5,000 genes — that’s a lot of variables. Fortunately, her collaborator from Harvard had identified a genetically tractable choline consumer and knocked out the TMA production gene. “Now the only difference in my communities was a single gene.”
Cardiovascular risk aside, choline is an underappreciated nutrient contributing to the process of epigenetic regulation of gene expression, and those without enough of it are more likely to suffer metabolic disease. In mice and rats, there is even a two-day window during pregnancy where lack of choline can impact fetal brain development. “Biology is never simple,” says Romano. “If it’s simple, you’re missing something.”
Eat Your Vegetables
Talking about poop makes people laugh, and as Rey wraps up in Reedsburg, the crowd has stayed engaged, surviving even his brief foray into 16S sequencing.
His advice for microbial health is folksy and charming. “Spouses share more microbes than people that don’t live together,” Rey tells the crowd. “We have found that spouses who get along together share more microbes than spouses that don’t. There is a lot of exchange going on there.” (In fact, if you’ve had to take antibiotics, he suggests family time will do more to restore your microbes than probiotics.)
And, noting that the most diverse microbiomes are found in places like the Amazon, he says being exposed to dirt is probably a good thing. “Get your hands dirty working in your garden,” he says. “I think that’s a health habit that we have lost over the years.”
Still, stubborn ideas persist among those gathered in the room. About a dozen times people bring up their pet microbiome theories for validation: probiotics, kombucha, fermented food, raw food, red wine vinegar, minimal vegetable washing, fecal transplants.
“The microbiome has come to mean anything you want it to mean,” Rey says disarmingly, for another laugh.
But “I don’t know” is his most honest answer. It’s a conundrum: The microbiome is hot in part because of some stunning findings. Most remarkable is the use of fecal transplants to cure drug-resistant Clostridium difficile infections, with cure rates running above 90 percent in some studies.
That extraordinary outcome certainly got the attention of both the medical community and the fad diet community. And even as it validates the power of the microbiome, that outcome actually runs against the grain of all the variation Rey is trying to figure out. “My lab is very interested in understanding the consequences of our interpersonal differences,” he explains.
“I can sequence your microbes, and I would not be able to tell you what vegetables to eat,” he says. “Maybe in 5 or 10 years personalized nutrition will be a reality, but it is not today. The one recommendation we can give right now is try to think about feeding your microbes. Because we cannot tell you what will be the best for your microbiota.”
In other words, eat your vegetables. Let’s say you eat pizza, with regular flour dough and cheese. Your body can digest every single ingredient of that pizza. By the time it reaches your large intestine, where most of your microbes live, your body has absorbed everything of nutritional value.
“You’re not sharing any of your food with your microbes,” he explains. “That’s one of the things we are doing with our Westernized diets: we are starving our microbes.”
“Maybe broccoli is the best for your microbiota whereas cabbage is the best for my microbiota,” he concludes. “But in general, if you eat a diverse diet that contains plant polysaccharides, eventually you are going to help the good guys